Analytics & Measurement

Marketers lose confidence as AI floods measurement with conflicting data

Marketers lose confidence as AI floods measurement with conflicting data FAYFO Media © fayfo.com
Marketers lose confidence as AI floods measurement with conflicting data © fayfo.com
Marketers face a crisis of confidence as more data and AI tools create confusion instead of clarity. New research shows trust in measurement is falling fast.

Senior marketers are watching their trust in numbers slip away. Dashboards multiply, but certainty vanishes. AI was supposed to bring order. It hasn't. Instead, the flood of new signals and tools has left teams more confused about what actually drives results.

Kantar's latest research, previewed at The Drum Live, shows a sharp drop in marketers' belief that they can get their media mix right or use their data well. The fall has sped up in the past year. Two things are fueling it: endless platform splits and the rush to adopt AI. The study surveyed 800 senior marketers worldwide. The finding is blunt. More data means less confidence in which metrics to trust.

LiveRamp and the MMA found that when identity precision drops to 50%, measured campaign return can fall from $1.50 to just $0.43, highlighting how fragmented identity distorts ROI.

PPC Land

One panelist called it a paradox. Every new platform promises better attribution. Retail media brings transaction data closer to ad exposure. But the outcome is a mess of clashing numbers. Gonca Bubani, global media director at Kantar, said the sheer volume of data is causing panic. Marketers don't know what to focus on. They struggle to combine data from different sources.

Another study cited at the event found that 97% of executives think the amount of data now outpaces their ability to connect and interpret it. Reza Amiri-Garroussi, head of media at LiveRamp, put it simply. More data just means more versions of the truth. In a fragmented ecosystem, platforms can grade their own homework.

AI is being sold as the fix for this chaos. The panel disagreed. AI can't make up for bad or inconsistent data. If you feed AI mismatched audience definitions or weak measurement frameworks, you get faster answers that are no more accurate. Amiri-Garroussi was clear. AI is only as good as the data it gets. Without shared definitions and agreement on what counts as success, automation just speeds up the confusion.

The advertising identity stack remains highly fragmented, with solutions like LiveRamp’s RampID, Unified ID 2.0, ID5, Yahoo ConnectID, TransUnion’s TruAudience, and PayPal Ads identifier all coexisting-yet none has emerged as a universal standard.

PPC Land

Some marketers are now asking if the industry is measuring too much just because it can. Last-touch attribution took heat for giving too much credit to the final customer interaction. It ignores the combined effect of channels like CTV, social, and the wider web. Click-through rates and impressions are easy to report. But they rarely show the real business impact of advertising.

The panelists agreed on one thing. The industry needs to move toward incrementality and outcome-based measurement. That means linking media exposure directly to changes in sales and business results. But this shift is not just about tech. It's an organizational problem. Visible metrics are popular because they help teams defend their budgets. It's hard to move away from them until everyone agrees on what success looks like.

Mike McCoy, president at Carat, cut to the core. The problem isn't the measurements. It's whether organizations have a clear strategic truth and can get teams to rally around it. Without that, measurement frameworks will keep splintering. AI will only make the noise louder.

The fight between short-term performance and long-term brand building is still unresolved. Kantar's research found confidence in balancing these goals has dropped. Last year, about 60% of marketers felt they had the mix right. This year, it's down to 50%. Only 40% say they have a reasonably balanced mix of short- and long-term goals.

Getting ready for AI-driven measurement takes groundwork. Teams need to recatalog and relabel assets and datasets so everyone can understand them the same way. This job needs C-suite ownership and coordination across channels. Isolated team efforts won't cut it. As shown in recent reporting, even unified dashboards can't solve the problem if the data underneath is still fragmented.

AI might eventually streamline MMM, attribution, and incrementality testing. Not yet. For now, its arrival exposes a deeper problem. Marketers don't need more data sources. They need agreement on which data matters. Until organizations set shared definitions and frameworks, the flood of AI tools will only speed up the race to nowhere. The obsession with visible metrics and fast reporting is hurting the industry's ability to measure what counts. Marketers who want real clarity have to do the hard work. Align on outcomes first. Only then should they chase the next wave of automation.

Ken Doctor Media analyst FAYFO Media
Media Analyst

Ken Doctor

An American media analyst, journalist, and publishing strategist